This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. The laboratory of Dr. Mario Capecchi at the University of Utah's Eccles Institute of Human Genetics is investigating the phenotypic expression of specific, induced genetic abnormalities in mice, a model that has been shown to provide insight into the ontogeny of congenital human disease. Conventional analysis of mouse skeletal structure requires sacrificing the research animal and a labor-intensive, time-consuming process of skeleton preparation and physical inspection under a dissecting microscope. Many tens or even hundreds of specimens are often required for a meaningful statistical analysis, which represents an enormous investment of time and money. The goal of the Center for Integrative Biomedical Computing collaboration with the Capecchi lab is to develop a faster, non-invasive protocol for skeletal analysis that uses semi-automated image processing of three-dimensional micro-CT rather than hand measurements of prepared skeletal specimens. We are developing a set of image segmentation, measurement and visualization tools for quantitative morphometry that allow us to experiment with new metrics such as the analysis of bone shape that would not be possible with prepared skeletal specimens. Furthermore, we expect that our tools will allow for more precise and repeatable measurements for length, density and volume, and therefore give insight into genetic alterations that have previously been described as pleiotropic (partially penetrant) or that have been misinterpreted as minor effects. In the short term, the Center for Integrative Biomedical Computing is targeting two specific research projects for publication. The first project is to validate our non-invasive CT-based protocol for skeletal analysis against the results obtained using prepared specimens and manual bone measurements by researchers in the Capecchi lab (Boulet and Capecchi, 2002;Davies and Capecchi, 1994). In this study, we will use scalar measurements of bone length and bone taken with our image processing and visualization tools. As in the Boulet-Capecchi study (Boulet and Capecchi, 2002), the length of the various bones of the paw will be compared to the length of the humerus. Our hypothesis is that we can reproduce the physical measurements to a greater accuracy (smaller standard deviation) and perhaps even measure additional variation that was lost in the measurement noise inherent to the physical study. Our second research project will apply our methods for computing statistical shape models to the segmented mouse bones. The mouse bones are a very challenging data set because their surfaces are composed of many complex and irregular features. We have developed a new technique for computing shape correspondence points, an essential step in the shape analysis pipeline, that we believe are more suited for these surfaces than conventional methods which parameterize surfaces as spheres. REFERENCES "Duplication of the Hoxd11 Gene Causes Alterations in the Axial and Appendicular Skeleton of the Mouse", Anne Boulet and Mario Capecchi. Developmental Biology, 249, 96-102, 2002 "Axial homeosis and appendicular skeleton defects in mice with a targeted disruption of hoxd-11", Allan Peter Davies and Mario Capecchi. Development, 120, 2187-2198, 1994.